© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Jacobian LensNEW
    Natural Language
    Autoencoders
    NEW
    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsinterp-engineNEWAPI Community BlogPrivacy & TermsContact
    1. Home
    2. Qwen3-4B
    3. 31-TRANSCODER-HP
    4. 18520
    Prev
    Next
    INDEX
    Explanations

    math problems

    np_max-act-logits · gemini-2.0-flash
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    mwhanna/qwen3-4b-transcoders/layer_31.safetensors
    Prompts (Dashboard)
    16,384 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    163,840
    Data Type
    float32
    Hook Name
    blocks.31.mlp.hook_in
    Architecture
    transcoder
    Context Size
    8,192
    Dataset
    monology/pile-uncopyrighted
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
     Eight
    -0.32
    åħ«åįģ
    -0.32
    åħ«
    -0.31
    8
    -0.30
     eight
    -0.30
    诫
    -0.29
    Eight
    -0.28
    ï¼ĺ
    -0.28
    ighth
    -0.28
    åħ«ä¸ª
    -0.28
    POSITIVE LOGITS
    åįģäºĮæĿ¡
    0.29
    çľŁäºº
    0.25
     turnovers
    0.24
    çĦ¦
    0.24
    .Serial
    0.24
    rans
    0.24
     instance
    0.23
    olle
    0.23
     Bundy
    0.23
     ham
    0.23
    Activations Density 5.598%

    No Known Activations